Genomic imprinting is a reversible and differential mark on a set of chromosomes that is determined by the sex of the transmitting parent. This imprint may dramatically affect chromosome behavior or gene expression, and disruption of a proper imprint may result in chromosome loss, cancer, birth defects, or genetic disease. An individual has both maternal and paternal imprints, reflecting that half of its genome has been inherited from each parent. Both sets of imprints are stably maintained throughout the life of an organism. In some cells (e.g., human sperm stem cells), this maintenance may last through thousands of cell divisions over the course of a century. As each chromosome is transmitted to the next generation, all parental imprints are erased in favor of a new set of imprints appropriate to the sex of the transmitting individual. However, since the imprints are maintained, established, and interpreted through unknown mechanisms, it is not known how loss of imprinting leads to disease. Our long-term goal is to understand how sex-specific imprints are established by both males and females, and maintained by the offspring, resulting in parent-of-origin-specific genetic behaviors. Our specific hypothesis is that DNA methylation is an important but transient feature of genomic imprinting. We base this hypothesis on analysis of genomic imprints in fruit flies mutant for the sole known DNA methyltransferase, Mt2. Our preliminary data suggest that DNA methylation is required maternally for establishment of paternal imprints, indicating that the paternal imprint is controlled by the maternal genotype. Based on these observations, we propose to focus our research on imprint establishment by pursuing three specific aims: 1. Identify the DNA sequence that is targeted for methylation by Mt2. We will identify which sequences become methylated by introducing an imprintable transgene, active throughout development and post-mitotic adulthood, and monitoring DNA methylation levels and chromatin structure as they correlate with gene activity. 2. Determine the mode of specificity of Mt2activity for paternally-derived chromosomes. Thekey feature of genomic imprinting is the discrimination of paternal from maternal chromosomes, without regard to chromosome sequence. We will focus on testing possible models of chromosome discrimination, including localized gene activity and differences in chromatin structure of maternal and paternal genomes. 3. Identify genetic factors responsible for establishing genomic imprints. We have identified a genetic interval containing a gene required for the establishment of a maternal imprint. We will identify which gene is involved in maternal imprint establishment, and begin to investigate its mode of action.